342 CALVIN B. BRIDGES 



heterozygous for cream II : 1 eosin pure for cream II. The entire 

 ratio, disregarding sex, approximated 12 : 1 : 2 : 1, the 12 being 

 the red-eyed flies. 



From the F2 a few cream males were selected and bred to their 

 sisters, all of which were wild-tj^pe in appearance, though a quarter 

 of them were homozygous for the cream gene (not-eosin creams). 

 This mass culture gave the expected cream females and males 

 from which a pure breeding stock was made up. There was a 

 difference in the color of the males and females of this pure stock, 

 the difference being of the same order as the normal bicolorism of 

 eosin. A complete separation of the eosin from the eosin heter- 

 ozygous for cream had not been attempted in the original F2 

 culture. In order to observe the heterozj^gous condition more 

 closely, a cream male from the pure stock was outcrossed to an 

 eosin female. The Fifties, both males and females (culture M688) 

 intermediate males 73, intermediate females 88,^ were lighter in 

 eye color than standard eosin, though the difference between 

 eosin and these heterozygotes was less than the difference be- 

 tween the heterozygotes and the pure cream. 



Among these Fo offspring (table 2) there were six different eye 

 colors: among the males the same three that occurred in the 

 original Fo and among the females three colors which, though cor- 

 responding genetically to the classes among the males, were darker 

 in eye color. The cream female is lighter than the normal eosin 

 male, while the heterozygous cream female is somewhat darker 

 than the eosin male. In order from the darkest (a deep slightly 

 yellowish pink) to the lightest (a pale translucent yellow) the six 



® One of the 88 intermediate daughters had only three segments to her abdo- 

 men instead of the usual five. This female (figured bj^ Morgan, '15, p. 425, text 

 figure 3a) was the founder of a new tj'pe of abnormal segmentation of the abdomen 

 — 'patched.' The segments were reduced in number (as in the first specimen) or 

 more typically were cut sharply into oblique or triangular pieces which were 

 patched together as illustrated in Morgan's figures b-f . This character was reces- 

 sive, but it generally reappeared in very much less than a quarter of the Ft off- 

 spring. The usual causes for such deficiencies are poor viability, partial or com- 

 plete dependence for realization on the coaction of one or more other genes, or 

 failure to be developed in all the flies genetically pure for the gene, whether from 

 environmental differences or because the normal fluctuations of the character 

 overlap the wild-type. 



